In recent years, a number of mtDNA mutations have been identified that cause maternally-inherited diseases, the majority of which are associated with mental retardation in early childhood. We propose to focus on two such disorders, both associated with mtDNA point mutations in polypeptide coding genes: mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes (MELAS), and maternally-inherited Leigh syndrome (MILS) (which is genetically related to another disease, neuropathy, ataxia, and retinitis pigmentosa, or NARP). MELAS is due to a mtDNA mutation at nt-9957 in subunit III of the cytochrome c oxidase gene; MILS/NARP is due to a mutation at nt-8993 in subunit 6 of the ATP synthetase gene. Focusing on these 2 mutations, we propose to create cellular and animal models of mitochondrial disease, using three related strategies: (l) Cybrid models: We will transfer patient mitochondria harboring these mutations to human p/o cells containing no endogenous mtDNA, thereby creating p/o-cytoplasmic hybrids (cybrids). This will enable us to study the relationship between genotype and phenotype in a neutral nuclear/mitochondrial background, and will help clarify the pathogenesis of these disorders, which is obscure at present. (2) Cellular models: We will begin to address the possibility of a genetic approach to treatment of these fatal disorders, by "recoding" the ATPase 6 gene (in the case of MILS~NARP) and COX III gene (in the case of MELAS) to contain the universal genetic code by in vitro mutagenesis, adding a mitochondrial targeting sequence, and transferring this construct to the nuclear genome. Expression of a recoded wild-type gene should ameliorate the effects of the respective mutations in mutant cells; expression of a mutant gene should create a dominant-negative phenotype in wild-type cells. We will also mutate bacterial ATPase 6 and COX Ill to mimic the disorder in a simpler, more-easily-manipulatable system. (3) Animal models: The same nuclear recoding-and-retargeting concept will be applied to create transgenic mouse models of NARP and MELAS, which would be the first animal models of mitochondrial disease.